Growth requirements and immunophenotype of acute lymphoblastic leukemia progenitors

Cox et al¹  reported that interleukin-3 (IL-3), IL-7, and stem-cell factor (SCF) promote the expansion of B-lineage acute lymphoblastic leukemia (ALL) cells. We cultured 9 cases of B-lineage ALL (8 CD10⁺ CD19⁺, 1 CD10^-CD19⁺) with IL-3, IL-7, and SCF as described by Cox et al. Cells became apoptotic, and viable leukemic-cell recovery after 2 weeks of culture was less than 0.01% to 3.6% (median 0.2%) of the input cell number. Cox et al¹  also reported the complete failure of ALL cells to survive on mesenchymal cells in their study, even in the presence of the factors that they found to be supportive. In 5 B-lineage ALL cases, we compared the cultures with growth factors described by Cox et al¹  to cultures with mesenchymal cells (without growth factors or serum). Median leukemia-cell recovery after 1 week of culture with growth factors and no mesenchymal cells was only 5% (range 0.01%-45%) of input cells; after culture on mesenchymal cells, it was 129% (range 55%-176%). After 3 weeks, cell recovery was 1% in all 5 cultures with growth factors. By contrast, viable leukemic cells persisted in 4 of the 5 cases cultured on mesenchymal cells, representing 17%, 25%, 108%, and 147% of the cells originally seeded. These results contrast with those reported by Cox et al,¹  but are in agreement with those of previous studies indicating that mesenchymal cells are essential to support ALL cell viability in vitro,^2-8  whereas growth factors may induce short-term proliferation in a few cases of ALL but do not support viability and expansion of ALL cells.^4,9 ^-11 

A second issue raised by Cox et al¹  is the immunophenotype of the putative ALL stem cells. We think that there are technical issues that must be addressed before concluding that the only cells that can engraft immunodeficent mice lack CD19 (or CD10) expression. Cox et al¹  used an anti–CD19 antibody conjugated to fluorescein isothiocyanate, which, as shown in Figure 1 of their paper, yields a relatively weak staining resulting in relatively high percentages of CD34⁺CD19^- cells. We reviewed expression of CD34 and of CD19 (detected with allophycocyanin- or phycoerythrin-conjugated antibodies) in 50 B-lineage ALL samples: with few exceptions, CD34⁺CD19^- cells were either undetectable or formed a separate cluster, apparently unrelated to the bulk leukemia-cell population. Hotfilder et al¹²  sorted TEL-AML1⁺ leukemia cells according to their CD19 expression in 8 cases and found that CD19^- cells either lacked TEL-AML1 transcripts (n = 3) or contained an estimated percentage of TEL-AML1⁺ cells ranging from 0.1% to 6% (n = 5). The latter percentages correlated with those of contaminating CD19⁺ cells in control sorts, suggesting that the low TEL-AML1 signals observed in the sorted CD34⁺CD19^- cells could have been derived from contaminating CD19⁺ cells. These authors did not perform ALL cell-growth assays with CD34⁺CD19^- cells, but observed that they gave rise to myeloid, monocytic, and erythroid colonies, indicating that most, if not all, were normal hematopoietic cells. In fact, Anthony et al¹³  recently reported that expression of CD19 was useful to distinguish leukemic CD34⁺CD38^- cells from normal CD34⁺CD38^- primitive hematopoietic cells in B-lineage ALL samples. It may be that CD19 expression is weaker on ALL cells with the least stringent growth requirements, or that CD19 on these cells modulates more actively after antibody binding, but it is premature to conclude that ALL stem cells lack CD19. This issue needs careful attention because CD19 is a candidate target for immunotherapies of B-cell malignancies, including B-lineage ALL.